"Copenhagen Consensus" theme:
cost-effectiveness versus cost-benefit
    Home                                                                                                    Copenhagen Consensus main page

Lomborg stresses that the money must be allocated to those projects where they do most good. This idea is, of course, not new. For many years it has been common practice to calculate the cost-effectiveness of interventions to treat or prevent diseases. It has been less common to calculate the cost-benefit ratio. Let us have a look at the difference between the two approaches.

Let us first look at cost-effectiveness. If we want to prevent the spread of HIV/AIDS, it is very effective to distribute free condoms. This will on average prevent one HIV infection for a cost of about US$ 14. You may also treat HIV-infected adults with anti-retroviral drugs to avert an outbreak of AIDS. This will probably cost about US$ 100,000 per outbreak averted, depending on the price you have to pay for the drug. So if you want to prevent that people die from HIV/AIDS, it is about 7,000 times more effective to distribute free condoms than to treat people with drugs once they are infected. Such figures are useful when you want to prioritize your efforts. If you have a limited funding for fighting the disease, then it is sensible to launch a full-scale effort for that type of intervention which is most cost-effective. That done, you may proceed with that intervention which is next most effective, and so on until all your money has been spent. By following that principle, you will more or less have obtained an optimal prioritization of your money.

Cost-effectiveness figures can of course also serve as guidelines when you are forced to choose between efforts against different diseases. For instance, if treatment of one case of tuberculosis costs US$ 500, and treatment of one case of pneumonia costs US$ 2, you will help more people by allocating your resources to pneumonia treatment.

One may argue that it is more important to save the lives of young people than the lives of old people who would soon have died anyhow. Therefore it is common to speak of "years of life lost" (YLL). If for instance the average lifetime of a person in a society is 67 years, and an individual person becomes deadly ill at an age of 44 years, then the years of life lost are 67 - 44 = 23. If your ambition is to save as many life-years as possible, then you may calculate the costs relative to the number of YLLs spared. Rather than simply counting the years, it is common to count the "disability adjusted life years" (DALY), i.e. a figure where years with reduced life quality and reduced ability to work do not count as full years. It is usual to calculate the cost effectiveness as costs per DALY.

All this works well with the cost-effectiveness measure. What is the difference between this and calculating the cost/benefit ratio ?

The principle in the cost/benefit method is that not only the costs, but also the benefits are expressed in money units. That is, in order to do this calculation, you must put a price on a human life or on a DALY. The usual method to do that among economists is to calculate the value of a year of life lost from the per capita Gross National Income in the country concerned. For example, in a paper (by Richard S. Tol (2002): Environment and resource economics 21: 47-73) life loss is valued at 200 times the per capita income. Once the price of a human life has thus been decided upon, it is rather easy to go from cost-effectiveness to cost-benefit calculations.

What is unpleasant about this, is of course that it is not nice to express the value of a human life in monetary terms. It is especially unpleasant if the per capita income is, say, 100 times higher in a very rich country than in a very poor country, because that means for instance that the value of a human life in Mozambique is worth only 1/100 of a human life in Switzerland. When calculating what pays on a worldwide scale, one will nearly always find that saving lives in Mozambique is nearly not worthwhile, as long as it happens at the expense of moderate economic disadvantages in rich countries. This is a good reason why many persons altogether reject the principle of cost/benefit analysis as it is used by Lomborg.

But why, then, does Lomborg insist on using the cost/benefit principle? It cannot be because of the need to prioritize the money, because prioritization is certainly possible on the basis of cost-effectiveness calculations. Lomborg´s reason is that cost/benefit allows prioritization of money that could be allocated to widely different economic sectors. He wants to compare sectors as different as human health, climate change, armed conflicts and financial stability. In order to do that, all kinds of benefits must be expressed in the same units - money units.

This goes against common thinking. There are many values that cannot be expressed in monetary units. For instance, the preservation of plant and animal species is seen by many people as an ethical obligation independent of the eventual monetary value of a species, just as the preservation of a human life is seen as an ethical obligation, even if that human person does not contribute to the gross national income. This way of thinking does not go against wise use of money - it is possible to calculate cost effectiveness both when saving species from extinction and when saving life years for elderly people.

Another kind of non-monetary value is the value of security, peace and harmony. For instance, if an elderly lady walks in the street, and a young boy comes running by and snatches her bag, then the loss is not equal to the monetary worth of the stolen bag. The loss also includes the loss of security - the lady does no longer dare to walk in the street, and the induced fear cannot be evaluated in money terms. If a society has a small difference between rich and poor, then the society will be relatively harmonious, and people will tend to be more happy. But this happiness cannot be expressed as a sum of money.

Therefore, by insisting on comparing anything and expressing all benefits in monetary units, Lomborg goes against the conviction of many people, and he should not expect that people would accept the results of such calculations.

This is not the only objection to the cost/benefit method. Another important objection is that you cannot or should not compare widely different benefits that accrue within widely different time horizons. The method used by Lomborg and many economists to do just that, is the method of discounting. By choosing a discount rate, one chooses how much a benefit in a far future is worth relative to a benefit that accrues already next year. But such a judgment cannot be made in an objective way; it will always be subjective, and the choice of discount rate is subjective. If my time preferences are different from Lomborg´s time preferences, I will use a different discount rate (or no discount rate at all), and I will arrive at a cost/benefit ratio very different from the one Lomborg arrives at. Furthermore, when the discount rate is set at a positive value, we presuppose that future generations will be richer than we are, and that a certain good which is important to us today will be less important to them, relative to their greater wealth. But we cannot know if this presupposition is true. It could be that certain goods will be extremely valuable to future generations, and that they should therefore be discounted with a very small or even negative discount rate, whereas they will not care at all about certain other goods which could therefore be discounted with a high discount rate. Only he who knows what will happen in the distant future - and nobody knows that - can tell what discount rate will be correct. Therefore it is absurd to postulate that we can make sensible choices of discount rates now.

Finally, one further objection is that the cost/benefit method neglects the precautionary principle - another point where values cannot very well be expressed in monetary terms.